Bath · Bristol · London
The Crane Index · Newsroom

Press and downloads.

Everything to run the story. The Crane Index measures how cleanly a machine can read and represent the leading organisations in an industry. Every figure below is a first-party measurement, free to quote and republish with attribution, and re-runnable live in about a minute.

Headline findings

  • The FTSE 100 (measured 1 September 2026). Of Britain's largest companies, 79 were approached and 40 could be read by a machine at all, and not one reached the Strong band of 70 out of 100. The field is led by GSK at 64. Budget bought none of them legibility.
  • The UK's biggest online retailers (measured 1 September 2026). Of 142 major storefronts, only 77 could be read by a machine, and retail's identity read collapsed to a median of 28. The list is topped not by a giant but by Office at 68.
  • The same shape in every industry (measured 1 September 2026). The technical reads hold up and the read that tells a machine who you are does not. Across all 20 benchmarks the median Agent Readiness score runs between 62 and 69 out of 100, while median Brand Representation runs between 13 and 33, and not one industry has it the other way round. That gap, rather than any single score, is the story.
  • The programme. Twenty benchmarks: the FTSE 100, the UK's biggest online retailers and 18 industry and sector benchmarks. Across them we set out to read 3,336 of the leading organisations in their fields and could read 2,291, each benchmark re-scored on the first of each month so the movement becomes public record. All 20 are listed below with their own figures.

The one-line story is the same in every cohort: the machine-readable web is early, structure decides rank rather than size or spend, and the top of every table is unclaimed territory.

Every benchmark, and what each one found

All 20 read to the same three-part method, so the medians are comparable across industries. If a figure for one industry was sent to you, this is where to check it and see the rest of the field around it. Each name opens that benchmark's own ranked table, where every organisation is listed by name and a site that could not be read is reported as unmeasured rather than scored.

Benchmark Read AI search Agent Brand Crane Index At 70+ Leads it for AI readability Take it
FTSE 100 40 of 79 48 64 14 43 0 GSK (64) PDF · CSV · JSON
UK Ecommerce 77 of 142 53 69 28 51 0 Office (68) PDF · CSV · JSON
Automotive 108 of 179 47 62 13 40 7 Vertu Motors (77) PDF · CSV · JSON
Consumer Goods 124 of 188 52 65 17 46 1 Levi Strauss & Co. (71) PDF · CSV · JSON
Education 92 of 123 53 68 13 47 4 Fettes College (75) PDF · CSV · JSON
Energy and Utilities 117 of 175 47 65 13 42 6 OVO Energy (78) PDF · CSV · JSON
Financial Services 119 of 175 52 63 15 44 6 Fidelity Investments (76) PDF · CSV · JSON
Healthcare and Pharmaceuticals 135 of 177 49 65 13 43 5 Hologic (76) PDF · CSV · JSON
Hospitality and Leisure 122 of 182 49 63 15 42 6 Orangetheory Fitness (79) PDF · CSV · JSON
Insurance 123 of 181 51 63 13 44 3 Everest Group (77) PDF · CSV · JSON
Legal Services 133 of 171 51 63 19 46 2 LegalZoom (79) PDF · CSV · JSON
Manufacturing and Industrial 132 of 176 53 66 19 46 4 Konecranes (72) PDF · CSV · JSON
Media and Entertainment 124 of 172 47 63 13 42 3 Nintendo (76) PDF · CSV · JSON
Professional Services 145 of 172 60 68 25 53 8 Mercer (76) PDF · CSV · JSON
Real Estate and Property 128 of 186 52 65 22 47 5 MRI Software (77) PDF · CSV · JSON
Technology and Software 123 of 169 64 65 33 57 17 Dynatrace (88) PDF · CSV · JSON
Telecommunications 134 of 184 49 63 15 43 7 Wessex Internet (78) PDF · CSV · JSON
Transport and Logistics 110 of 176 51 63 14 43 2 Agility (73) PDF · CSV · JSON
Travel and Tourism 88 of 171 55 63 23 47 3 Princess Cruises (73) PDF · CSV · JSON
Universities 117 of 158 51 68 13 43 4 University of Kent (88) PDF · CSV · JSON

Reading the columns. Read is how many of the cohort an automated reader could reach at all; the rest were approached and could not be read, and they are named on each table with the reason. Where a benchmark approached fewer organisations than its name suggests, its own page says why, as the FTSE 100 does. The three reads are AI Search Readiness (the retrieval read: can an engine resolve you and extract what you sell), Agent Readiness (the execution read: can an automated agent get in and act) and Brand Representation (can a machine confirm who you are). Crane Index is the composite of the three. Every median is the midpoint of the readable sites, so a site that refused a reader is left out of it rather than scored as zero, and At 70+ counts the sites in the Strong band. Leads it is the top-scoring public site, shown jointly where the top score is shared. Measured 1 September 2026, and re-measured on the first of each month. Every check behind these figures, and every exclusion, is set out in the methodology.

Any score in any of these tables can be re-measured against the live site in about a minute on the free Crane Index scanner, so nothing here has to be taken on trust. Every row's data is the same archive the tables are built from, as CSV for a spreadsheet and JSON for a chart, under the licence below.

A quote you may use

“The web spent thirty years being optimised for people. We have now read 2,291 of the leading organisations in their fields for how well a machine can understand them, and 93 of them score 70 or better. The Crane Index measures that gap in public, one industry at a time, and the striking part is that budget does not close it. Structure does.”

Nic Crane, founder, The Crane Consultancy

Figures measured 1 September 2026 across 20 benchmarks, and re-measured on the first of each month.

Downloads

Those are the programme-wide assets. Every one of the 20 benchmarks carries the same again in the table above: its own board report, and the archive the table is built from as CSV and JSON. Reports refresh on the first of each month. For the full method behind every figure, see the methodology.

Licence and how to cite

Released under a Creative Commons Attribution 4.0 licence: free to quote, chart and republish with attribution. Please cite as:

The Crane Index, the AI readability benchmark. The Crane Consultancy. thecraneconsultancy.com/research/

Press contact

For interviews, a specific industry cut, or a figure checked before you publish, write to [email protected]. We are glad to run a bespoke cohort for a story on request.

Every organisation in these tables is named, so every one of them has a right of reply. Any company can re-measure its own score live, ask for a re-scan if its site has changed, or have a domain corrected for the next monthly run. The methodology sets that out in full, alongside every exclusion.

Questions about the Crane Index

What does the Crane Index measure?

How readable and well-optimised a site is for AI, drawn live from its real pages across three deterministic reads: whether an AI engine can retrieve and cite it, whether an agent can act on it, and whether it is described accurately. Each is scored out of 100, and the composite is the Crane Index.

Why does AI readability matter now?

Because discovery is moving. In 2026, Google reported that its AI Mode in Search had passed a billion monthly users. What an assistant can read of an organisation comes down to the machine-readable layer of its site, the structured data Google Search Central documents.

How often is it re-run?

Every month, on the first, so the movement is a matter of public record rather than a single snapshot. Every constituent is named, ranked and archived, and each score can be re-run live against the site as it stands today.

The Crane Standard

The senior read,
once a week.

Most marketing newsletters are noise on a schedule. This is not that. Once a week I send a short, considered read on the shifts that actually change what you should do, across the AI shift, PPC, technical SEO and analytics. Filtered, weighted and signed off by me before it goes out. The judgement is mine, and now you can hear every edition in my own voice.

Nic CraneFounder, The Crane Consultancy

Topics you want

All three lands every Tuesday. Choose a selection and you get only those, as and when they publish, always on a Tuesday.

One considered email a week, edited and signed off by Nic. No filler, no sharing, unsubscribe in a click. By subscribing you agree to our Privacy Policy.

Prefer LinkedIn? Subscribe on LinkedIn instead, the same edition, delivered by push and email.